Not yet reviewed by Pith; the record is open.

The energy-dependent frame transformation theory of Gao and Greene 1990 [Phys. Rev. A {\bf 42}, 6946 (1990)] is extended to yield quantitatively accurate description of the dissociative recombination process. Evidence is presented to show that direct application of the original theory leads to inaccurate cross sections. A major revision, based on an interaction-free back-propagation of the Born-Oppenheimer solutions, markedly improves the frame transformation theory, reducing its average error by orders of magnitude. The original theory and its extension are tested on the previously explored 2D model that is tailored to describe the singlet ungerade states of molecular hydrogen. The 2D model can be solved exactly (within the numerical accuracy) without implementing the Born-Oppenheimer approximation. These exact results then serve as a benchmark for the frame transformation theory developed in this paper.